Nanoconfinement-Induced Phase Segregation of Binary Benzene

Subscriber access provided by UniSA Library

C: Surfaces, Interfaces, Porous Materials, and Catalysis

Nanoconfinement-Induced Phase Segregation of Binary Benzene-Cyclohexane Solutions within a Chemically-Inert Matrix Kathryn L. Krycka, Joseph A. Dura, Luther J Langston, and Christopher M Burba J. Phys. Chem. C, Just Accepted Manuscript • Publication Date (Web): 20 Mar 2018 Downloaded from http://pubs.acs.org on March 20, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Nanoconfinement-Induced Phase Segregation of Binary BenzeneCyclohexane Solutions within a Chemically-Inert Matrix Kathryn L. Krycka,1,* Joseph A. Dura,1 Luther J. Langston II,2 and Christopher M. Burba2,† 1

NIST Center for Neutron Research, National Institute of Standards and Technology, Gaithersburg, Maryland 20899 2

Department of Natural Sciences, Northeastern State University, Tahlequah, Oklahoma 74464

*Electronic address: [email protected] †Electronic address: [email protected]

Abstract Binary solutions provide a fertile arena to probe intermolecular and molecular/surface interactions under nanoconfinement. Here, the phase segregation of a solution comprised of 0.80 mole fraction benzene and 0.20 mole fraction cyclohexane confined within SiO2 nanopores was evaluated using small-angle neutron scattering with hydrogen-deuterium contrast matching. It is demonstrated that the benzene and cyclohexane are fully miscible at 303 K (30°C), yet they unambiguously phase segregate by 153 K (−120°C), which is below their respective freezing points and below the cubic-to-monoclinic phase transition of cyclohexane. Specifically, the cyclohexane and benzene separate into a core|shell morphology with cyclohexane concentrated toward the nanopore centers. Additionally, pure benzene is shown to form a frozen core of bulk density with a thin shell of slightly reduced density immediately adjacent to the SiO2 nanopore wall at 153 K. Since the SiO2 matrix is chemically inert to cyclohexane and benzene, the observed radially-dependent phase segregation is strong evidence for the effects of confinement alone, with minimal host-wall attraction.

ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Introduction Liquids confined within mesoporous materials often exhibit unusual properties, including reduced melting points and slowed molecular dynamics, and increased densities compared to unconfined liquids. The general consensus is that these effects arise from molecular interactions between the liquid and walls of the confinement host. In pores with relatively large diameters, liquids often form crystalline solids with a one or two monolayer thick non-freezable interfacial layer when the temperature is lower than the melting point of the confined liquid. In the small pore regime, however, crystallization may be completely suppressed.1 The physics of confinement-induced

melting point depression is typically described by the Gibbs-Thomson equation for which the change in the melting point of a confined liquid is inversely proportional to the radius of the pore. This approach adequately describes thermal properties for a large number of liquids that form strong intermolecular interactions with the wall, such as the archetypal example of the water-mesoporous SiO2 system. There are many examples, however, where the equation is deficient.2–7 Thus, understanding confinement-induced changes for liquids that lack strong interactions with the host wall, such as those with only van der Waals interactions like benzene4,6,8–10 and cyclohexane5, is especially important. While the phase behavior of confined, single-component liquids is fairly well understood, the properties of confined solutions remains comparatively unexplored. From both a fundamental research point of view and the potential applications involving confined liquids, it is vital to have a clear picture of how surface interactions influence the behavior of solutions. In particular, it is not known whether or not the non-freezable surface layer retains the same overall solution composition when confined solutions are frozen. Therefore, we have chosen to focus on a binary solution of cyclohexane and benzene to quantitatively assess the composition and morphology of


Page 2 of 25

Page 3 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


the non-freezable surface layer under confinement and in the absence of appreciable solutionwall attraction. Benzene and cyclohexane form non-ideal solutions at room temperature with positive deviations from Raoult’s law.11 The solid-liquid phase diagram has a single eutectic point at 234 K.12–15 Consequently, a frozen mixture of benzene and cyclohexane will contain a mixture of benzene and cyclohexane crystallites, with the relative amounts of each being governed by the overall sample composition. Although the compounds are hydrophobic, both solvents wet quartz surfaces, having sessile drop contact angles of ~10°.16 Moreover, the two compounds experience reduced melting point temperatures when confined within mesoporous silica, and the cubic-tomonoclinic solid phase transition for cyclohexane (186 K) also occurs at reduced temperatures under confinement (about 173 K).5 There is indirect evidence of a distinctive near-wall region in frozen benzene whose thickness, density, and ordering characteristics are under debate.7,9,10,17,18 Small-angle neutron scattering (SANS) is uniquely poised to address questions about how solutions interact with the pore walls of a mesoporous host.19 In particular, benzene, cyclohexane, and mesoporous silica have different scattering length densities (SLDs) that will enable contrast matching of the solution SLD to that of silica with protonated and deuterated versions of the solvents. This will provide a powerful set of tools for examining these solutions under confinement. Contrast matching a solution to the SLD of silica at room temperature will minimize the intensities of the Bragg peaks associated with the mesopore lattice. If the nonfreezable layer retains the overall solution composition upon freezing, the average SLD−apart from small density changes−should remain contrast matched to silica. Conversely, compositional variation of the non-freezable layer will break the contrast matching between the solution and the pore walls, giving rise to Bragg peaks of measurable intensity.



SANS has previously been used to examine the interaction of benzene in confined silica samples. Early work by Ramsay and Hoinkis20 used SANS to explore adsorption of deuterated benzene on porous silica gels, while Xia et al.17 used neutron scattering to investigate the solidliquid phase transition of benzene confined within MCM-41 (2.4 ≤ d ≤ 3.6 nm) and SBA-15 mesoporous silicas (4.7 ≤ d ≤ 14 nm). Static structure factors obtained from neutron diffraction of the frozen samples (T = 70 K) revealed crystallization of benzene within the larger-sized pores, but benzene vitrified in smaller pores to form an amorphous solid. Lin et al. applied contrast matching to examine liquid-liquid phase separation in the H2O-2,6-lutidine system.21 Hellweg et al.22 and Schemmel et al.23 also used SANS to demonstrate de-mixing of confined binary solutions of iso-butyric acid and D2O below the upper critical solution temperature. In a similar set of SANS experiments, Hamid and co-workers24 established microphase separation of iso-butanol and toluene upon confinement. Iso-butanol forms hydrogen bonds with Si-O and SiOH groups along the pore wall-solution interface; thus, this component accumulates along the pore wall to produce an iso-butanol-rich shell with a toluene-rich core at room temperature. All of these systems differ from our work in that in those cases, one or both of the components is capable of hydrogen bonding with silica. Neither benzene nor cyclohexane is able to form hydrogen bonds; thus, we do not anticipate either compound to have a particularly strong affinity for the pore wall. Experimental Methods Our sample involves a mesoporous silica (SiO2) matrix of hexagonally-packed cylindrical pores purchased from Sigma Aldrich. Characterization of the mesoporous silica is provided in the Supplemental Information. Notably, it appears to contain a large fraction of voids by volume, which are solution penetrable. Differential scanning calorimetry (Mettler DSC 1) was used to


Page 4 of 25

Page 5 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


identify thermal transitions from mesoporous silica flooded with benzene-cyclohexane solution. DSC samples consist of 4.3±0.6 mg mesoporous silica mixed with 24.9±3.3 mg solution, yielding a silica mass fraction of 14.9±2.0%. Samples were then hermetically sealed in 40 µL aluminum crucibles, and thermal transitions recorded at a 10 K min-1 scan rate under a dry nitrogen atmosphere. SANS patterns were collected at the NIST Center for Neutron Research beam line NGB 30 m SANS. Samples consisted of mesoporous silica packed into a 1.0 mm thick sample holder with quartz windows. The sample cells were filled with mesoporous silica and either flooded with excess benzene/cyclohexane solution (0.80 mole fraction benzene), flooded with pure benzene, or left open and baked, before sealing. The incident neutron wavelength was 0.60 nm with an 11% full-width half maximum wavelength spread. Data were collected with a 2D detector at sample-to-detector distances of 13.035 m, 4.535 m, and 1.865 m, with the detector offset by 25 cm in a direction perpendicular to the incident neutron beam at 1.865 m in order to obtain the highest . The data from the three detector distances were combined and placed onto an absolute scale using in-house software.25 Results and Discussion Description of the Mesoporous Silica As described above, the pores of the mesoporous silica are either unfilled (Open) or filled with deuterated benzene (DB), 0.80 mole fraction deuterated benzene and 0.20 mole fraction deuterated cyclohexane (DBDC), or 0.80 deuterated benzene and 0.20 protonated cyclohexane by mole fraction (DBPC). This particular composition of benzene is selected such that the DBPC sample will be approximately contrast matched to the SiO2 host matrix as viewed by neutron



scattering when the solution is fully miscible. The pores are hexagonally close packed, Fig. 1, as indicated by transmission electron microscopy (TEM). The pore-to-pore distance, , shows some variation between 11.5 nm (blue marker in Fig. 1b) and 11.9 nm (red marker in Fig. 1b). X-ray scattering places the primary inter-particle peak at 0.057 Å-1 (see Supplemental Information). In a hexagonally close-packed system this would correspond to the √3/2 reflection, yielding

Figure 1. (a-b) TEM images indicate the pores are hexagonally close-packed. (c) TEM crosssection of the pores indicates they are relatively straight and thin walled. (d) SANS set-up (not drawn to scale) shows how an ensemble average of packed-pores over all orientations is probed in terms of angle .


Page 6 of 25

Page 7 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


= 12.7 nm. A TEM image from a slice along the pore walls, Fig. 1c, shows that the pores are relatively straight and thin walled deep within the sample interior. Comparison with the surface TEM image in Fig. 1b suggests that the pores narrow toward the surface. N2 physisorption data indicate that the internal diameter of the pores is about 9 nm in diameter, and Hg porisometry data indicates the mesoporous silica is highly porous. Phase Behavior of Confined Benzene-Cyclohexane Solutions Thermal transitions for benzene-cyclohexane solutions in the presence of mesoporous silica are presented in Fig. 2. The pure benzene sample exhibits two phase changes. The higher temperature phase transition is assigned to the fusion of unconfined benzene present in an external liquid reservoir. The lower temperature phase transition is attributed to the fusion of benzene that is confined within the silica mesopores. Four phase transitions are observed for the sample containing pure cyclohexane in contact with the mesoporous silica. Similar to pure benzene samples, the fusion of confined cyclohexane occurs at lower temperatures than unconfined cyclohexane. We also observe a reduction in the cubic-to-monoclinic solid-state phase transition temperature when cyclohexane is constricted within the mesoporous silica (190 K to 171 K); this finding is similar to other reports concerning the phase behavior of confined cyclohexane.5,8,14,17,26 Continuing to observe the 171 K phase transition, regardless of solution composition, provides compelling evidence that benzene-cyclohexane solutions phase separate within the pores at low temperatures. The phase transitions for unconfined solutions match the known phase diagram of benzene and cyclohexane.12 It should be noted, however, that the phase behavior of solutions containing deuterated components may be slightly different than the protonated analogs. Confined solutions, however, show relatively constant melting points at ~210 K over a broad range of solution



compositions (approximately 0.1 to 0.8 mole fraction benzene). Our samples consist of confined benzene-cyclohexane solution in contact with an external reservoir of unconfined solution. When the sample temperature falls below the freezing point of the unconfined solution, the majority component crystallizes in the external reservoir, and the unconfined solution composition will be driven towards the eutectic composition. It is plausible that confined solutions undergo mass exchange with the external reservoir and also adopt the eutectic composition. Hence, the 210 K phase transition may originate from the fusion of confined solutions that have the eutectic composition. These observations are similar to the findings of Meissner et al.27, wherein the melting point temperatures of concentrated aqueous salt solutions under confinement was reported to not exhibit concentration dependence. The 210 K phase transition is absent for dilute solutions on either end of the phase diagram in Fig. 2. Instead, the confined solutions appear to have composition-dependent melting points, which are inconsistent with mass exchange between an external reservoir that has adopted the eutectic composition. Similar behavior has been previously reported for dilute NaCl aqueous solutions confined within MCM-41 and SBA-15 mesoporous silicas.28 In this example, melting points of confined solutions depend on both solute concentration and pore size diameter. It is notable that fusion temperatures of dilute NaCl solutions confined within MCM-41 fall below


Page 8 of 25

Page 9 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 2. Phase transition temperatures for benzene-cyclohexane solutions in contact with mesoporous silica obtained from DSC (10 K min-1 scan rate). Transitions assigned to the confined solutions are marked with a dashed red line. the solidus temperature of the unconfined solution, suggesting these solutions are unable to achieve the eutectic composition. LiBF4 and LiPF6 dimethyl carbonate (DMC) solutions also experience concentration-dependent melting points when confined within MCM-41 and/or SBA15 silicas, regardless of the presence or absence of an external reservoir.28,29 For dilute solutions, such as those encountered at high and low mole fractions of benzene in Fig. 2, a substantial amount of solid phase will accumulate before the external reservoir freezes entirely. It is possible the excessive amounts of solid material will block the pore openings, thereby preventing the confined solution from undergoing mass exchange with the unconfined solution. Furthermore, thermal scanning rates as well as the relative volumes of the external



reservoir and the mesopores are two additional confounding variables that may play a role in facilitating mass exchange between confined and unconfined solutions during the freezing process. It should be stressed that all of these variables will likely affect the range of compositions over which the confined solution adopts the bulk eutectic composition. Our SANS experiments focus on solutions having 0.80 mole fraction benzene because this composition is almost perfectly contrast matched to the silica pore walls when prepared with deuterated benzene and protonated cyclohexane. This composition is very close to the cross-over point between composition-dependent and -independent melting point regions in Fig. 2. However, as shown below, the experimentally observed DBDC and DBPC intensity ratios for the first and second Bragg peaks in the SANS data are inconsistent with the confined solution having the eutectic composition. Instead, the SANS data are better explained by a confined solution that retains the nominal 0.80 mole fraction benzene. SANS of Confined Benzene and Cyclohexane Solutions Small-angle neutron scattering is ideally suited for measuring ensemble-averaged morphologies with dimensions on the order of nm to sub-microns. SANS probes the structure of our powder sample over all possible orientations in space. However, the measured intensity, I(q), associated with the ordered pores dominates when the momentum transfer, , is equal to 2 / , where is the spacing between planes of scattering centers and (the angle between the cylinder long axis and , Fig. 1d) is 90°. This scattering from periodically ordered structures is referred to as the structure factor, , which gives rise to the two obvious delta functions (Bragg peaks) in Fig. 3. The intensity of is weighted by the form factor of individual cylinders, , , discussed below. Additionally, a Porod30 (or power law) scattering of form


Page 10 of 25

Page 11 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


is observed when there is contrast between the surrounding medium and objects of sufficiently large size such that 2 /size ≪ . Here, this is attributed to the mesoporous silica grains in solution. Mathematically, the intensity is then expressed as:31,32 ∝ , " 2#

!

, ,

$%&'()/√* ()/√

+Δ-./01 2

(1)

34 (56789 / (5/

: Δ-;?9@@ / (5/

A

34 (56789 / (5/

:B

(2)

P is the Porod scale factor, and m is the exponent. F is the form factor for a core-shell cylinder31,32 of respective core and shell diameters, Dcore and Dshell, length L, cylinder volume #, and sample volume fraction CD . Δ- is difference in - from the SiO2 matrix. EF represents the first Bessel function. Instrumental smearing is calculated25,33 and broadens these delta functions into two prominent Lorentzian-shaped peaks, labeled Bragg1 and Bragg2 in Fig. 3.

Figure 3. (a) SANS scattering on absolute scale at 303 K (a) and 153 K (b). Except for the contrast-matched DBDC at 300K, the data are each fit with a Porod slope plus three Lorentzians at (0.058, 0.112, and 0.18) Å-1 of variable amplitude. Errors bars of one standard deviation are included here and everywhere else.



Page 12 of 25

- is a measure of how strongly a given material interacts with incident neutrons. By replacing naturally occurring, protonated cyclohexane (PC) with deuterated cyclohexane (DC), a markedly different - is obtained, Table 1. This control of contrast greatly enhances sensitivity to the placement of cyclohexane within the pore structures, which is observed as a change in the intensity of F2(q,D), Eqn. 2, for q values located at Bragg peak positions.

Table 1. Calculation of ρ as a function of miscible material compositions. The densities of PC, DC, and DB are 0.779, 0.893, 0.950) g cm-3, respectively, at 303 K and 0.996, 1.14, 1.21 g cm3, respectively, at 153 K. Compound

Composition

ρ at 303 K (10-6 Å-2)

ρ at 153 K (10-6 Å-2)

Mesoporous

SiO2

3.36

3.36

PC

C6H12

-0.279

-0.356

DC

C6D12

6.70

8.57

DB

C6D6

5.43

6.91

DBDC (nominal)

0.80C6D6 + 0.20C6D12

5.73

7.30

DBPC (nominal)

0.80C6D6 + 0.20C6H12

4.10

5.22

DBDC (eutectic)

0.22C6D6 + 0.78C6D12

N/A

8.26

DBPC (eutectic)

0.22C6D6 + 0.78C6H12

N/A

1.01

Silica

The experimental scattering patterns obtained at 303 K (liquid phase) and 153 K (solid phase) are shown in Fig. 3a,b. The solid curves correspond to the data fits in SasView32 involving a Porod slope plus three Lorentzians of variable intensity located at 0.058, 0.112,


Page 13 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


0.180 Å-1 with full-width half-maximums of 0.0042, 0.015, 0.050 Å-1, respectively, which capture the essence of Eqns. 1 and 2. Including Lorentzian peak width beyond the instrumental smearing is an indication that there is some polydispersity in the pore-to-pore distance. The parameters of these fits are given in Supplemental Table 1. For the 303 K solutions, a universal Porod exponent of m = 3.4 fits the data well. This value of m is quite reasonable for the mesoporous spheres since G " 4 corresponds to a perfectly smooth sphere, while G " 3 corresponds to a collapsed polymer coil.30 At 153 K, the Porod exponent increases slightly to 3.38 for DB, 3.46 for DBDC, and 3.60 for DBPC. These increases in m could indicate the formation of air-ice interfaces and/or the formation of cyclohexane-rich and benzene-rich frozen domains in the solution surrounding the mesoporous silica. The Open scattering is indistinguishable between 303 K and 153 K, confirming that no measurable thermal contraction of the SiO2 occurs upon cooling. Thus, the intensities of the Open Bragg peaks, denoted I0JKK,LM1N , are an ideal reference to which other data are compared. The relevant peak ratios of Supplemental Table 1 are provided here in Table 2, columns 1 and 2.

Table 2. Experimental and calculated

O>7@PQR7S O7T9S

"

'U>7@PQR7S UVRWX * 'YUVRWX *

X

X

ratios, where the uncertainty on

the experimental values is less than 10%. Calculated values assume that SiO2 ρ = 3.36 x 10-6 Å-2 and that the solutions are fully miscible, where miscible solutions intrinsically yield the same value for the Bragg1 ratio and Bragg2 ratio. .Sample DB at 303 K DBDC at 303 K DBPC at 303 K

Experimental Bragg1 ratio 0.40 0.47 Too small

Experimental Bragg2 ratio 0.45 0.50 Too small

Calculated, nominal ratio 0.38 0.50 0.049

Calculated, eutectic ratio N/A N/A N/A

DB at 153 K DBDC at 153 K DBPC at 153 K

1.17 2.13 3.24

0.67 1.20 0.91

1.16 1.38 0.306

N/A 2.13 0.489



Page 14 of 25

∝ Δ- for open pores and pores containing solutions that fully miscible. At 303 K, benzene and cyclohexane-benzene solutions meet these requirements, allowing for the determination of for the mesoporous silica matrix. Using the calculated -'s of DB and DBDC and matching the ratios of I0JKK4 for DB and DBDC to Open (Table 2), the best fit of - for SiO2 was determined to be 3.36 × 10] Å-2, which is only slightly less than the literature value for amorphous SiO2 at 3.47 × 10] Å-2.34,35 Due to near contrast matching between SiO2 and DCPC, which results in a low intensity Bragg1 peak at 303 K, the DBPC sample has been excluded from this fit. At 303 K, the close agreement between calculated and experimental

O_8àa4 ,>7@PQR7S O_8àa4 ,7T9S

ratios in Table 2, as well

as a similarity in Bragg1 and Bragg 2 ratios within ≈10% uncertainty, confirms that the solutions are miscible. In contrast, at 153 K the experimentally determined ratios of O_8àa4 ,b_dc O_8àa4 ,WT9S

O_8àa4 ,b_

,

O_8àa4 ,b_bc

O_8àa4 ,WT9S O_8àa4 ,WT9S

, and

cannot be explained by uniform, miscible solutions, Table 2. Notably, the measured

intensities of

O_8àa4 ,b_bc O_8àa4 ,WT9S

, and

O_8àa4 ,b_dc O_8àa4 ,WT9S

at 153 K are 1.54 and 10.6 times larger than what is

calculated for miscible solutions of the nominal composition, respectively. While the eutectic composition could explain

O_8àa4 ,b_bc O_8àa4 ,WT9S

, it cannot explain


. In fact, varying the

composition of the miscible solutions from 0% to 100% of cyclohexane cannot explain the O_8àa4 ,b_bc O_8àa4 ,WT9S

and


experimentally measured ratios. Moreover, the statistically different

Bragg1 and Bragg2 ratios per sample are a second indication that data cannot be explained by any


Page 15 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


miscible solution for which the Bragg peak ratios would be intrinsically the same (to be demonstrated next). The simplest, non-miscible morphology is that of a segregated core and shell running parallel to the cylinder-pore long axis (Eqn. 2). Modeled , scattering from miscible, cyclohexanecore|benzene-shell, and benzene-core|cyclohexane shell morphologies are shown in Fig. 4. Fig. 5a,b provides the corresponding solution to open pore scattering ratios. The use of a generic diameter, D, with the multiplication of q by √3/4 is described below. The most obvious feature of Fig. 4 and Fig. 5 is that miscible solutions produce ratios that are constant as a function of q (thick solid lines of Fig. 5) and, thus, they cannot explain the differing Bragg1 and Bragg2 ratios per solution observed at 153 K.

Figure 4. Simulated scattering for solutions of nominal composition at 153 K at several different possible morphologies. In a hexagonally close-packed structure, Fig. 1b, the lowest-q reflection peak would correspond to 4 /√3 where is the pore center-to-pore center distance. The lowest-q reflection of 0.058 Å-1 (or 0.057 Å-1 from x-ray diffraction, Supplemental Material) corresponds to = 12.5 nm. Due to porosity of the SiO2 matrix and likely widening of the pores away from the mesoporous silica surfaces, the exact outer pore diameter D (= Dshell) is unknown, but it must



Page 16 of 25

be ≥6.9 nm (see Fig. 1b) and ≤12.5 nm (). This means D could range from 0.55 up to 1.0 . However, altering D while keeping the ratio of core and shell volumes fixed simply stretches/compresses the scattering profile of , along q. Thus, a , model of generic D can be plotted against √3/4 , where the desired ratios of O_8àa4 ,b_bc O_8àa4 ,WT9S


≥ 3.24 and

≥ 2.13 (Table 2) must simultaneously occur at single value of √3/4 between

0.552 and 1.0. Additionally, the

O_8àaX ,b_dc O_8àaX ,WT9S

≥ 0.91 and

O_8àaX ,b_bc O_8àaX ,WT9S

≥ 1.20 (Table 2) must also

occur at single q-value of √3/4 between 1.1 and 2.0 that is approximately twice that of the first Bragg peak. These requirements are shown as dotted lines in Fig. 5 and Fig. 6. Note that ratios slightly different than experimentally observed can be valid because mixing of the cyclohexane and benzene or a graded interface between the core and shell tends to push the observed scattering closer toward the fully miscible case (solid thick lines of Fig. 5).


Page 17 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 5. Simulated scattering ratios for solutions of nominal composition at 153 K of miscible, benzene-core|cyclohexane-shell, and cyclohexane-core|benzene-shell morphologies. The dotted lines represent experimental data ratios, while the solid thick (flat) lines are the fully miscible models. The top and bottom rows correspond to nominal and eutectic solution compositions, respectively. The red circles indicate the best match among the models considered here, to be refined in subsequent text. Fig. 5 shows modeled , scattering ratios at 153 K for a variety of possible conditions: miscible solutions of nominal or eutectic composition (thick lines), core|shell segregation of solutions of nominal or eutectic composition with the benzene located at the core (column 1), and core|shell segregation of solutions of nominal or eutectic composition with the cyclohexane



located at the core (column 2). Neither nominal or eutectic compositions of miscible solutions (thick lines of Fig. 5) come close to reproducing the dotted lines required for the Bragg1 or Bragg2 ratios. The nominal and eutectic solutions with the benzene concentrated toward the pore centers also do not reproduce the dotted line requirements, Fig. 5a,c. However, concentrating the cyclohexane at the pore center, Fig. 5b,d can explain the experimental data. A significantly better match to both the Bragg1 and Bragg2 peak ratios is found for the solutions of nominal composition (Fig. 5b, red circles) than of the eutectic composition (Fig. 5d), since the eutectic Bragg2 ratios cannot meet experimental, dotted line values even with unlimited core and shell mixing. Some diffusion or gradient between the cyclohexane and benzene core|shell morphology is expected. Although there are many possible variations in modeling the boundary “fuzziness”, here a simple model is used that retains the core|shell boundary radius associated with a segregated 20% mole fraction cyclohexane and 80% mole fraction benzene solution (23.3% cyclohexane and 76.7% benzene by volume), yet the model allows a fraction of the cyclohexane and benzene in the core and shell regions within the pores to exchange while preserving the nominal composition within the entire pore. Setting the core to be 60% cyclohexane and 40% benzene by volume, the shell becomes 88% benzene and 12% cyclohexane by volume, as showed in Fig. 6a. Since the DSC data clearly indicate a first-order phase transition, the formation of a glassy solid is ruled out. Instead, the SANS and DSC data are consistent with the formation of benzene and cyclohexane crystallites in the interior of the pores. This is consistent with wide-angle neutron scattering experiments that show crystallization of benzene when confined within SBA-15 silica at 70 K.17


Page 18 of 25

Page 19 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Here, both Bragg1 and Bragg2 ratios are achieved for the DBDC and DBPC near √3/4 = 1 and 2, respectively. Other modeling variations of a graded surface also consistently require D to be close to κ, which implies that the internal pore diameters either widen to near 12.5 nm deep within the silica grains or that the silica is highly porous (consistent with the Hg porisometry results). Indeed, modeling reveals that porous silica volume fractions up to 25% (see Supplemental Material) can explain the data well; this is consistent with SBA-1536,37 and MSUH38 silicas where individual pores are believed to be interconnected with smaller nanopores. The key to a successful model is that the cyclohexane must be preferentially phase segregated toward the pore centers at 153 K.

Figure 6. Best-fit models where dotted line represent experimental Bragg peak ratios. (a) A cyclohexane-core|benzene-shell model at 153 K where solution exchange is allowed between the two regions. (b) Models of benzene at 153 K with reduced density shells, where a shell of 3.0 to 3.5 nm (1.75 nm of this shell within the porous SiO2) yields the best fits. SANS of Confined Benzene Finally, evaluation of a single component sample, containing only deuterated benzene, is of relevance in that it provides insight between competing models regarding the benzene distribution along the silica interface. While Xia et al.17 proposed a uniform increase in the



density of confined benzene throughout the pores, there are a class of models that suggest benzene may form a shell which differs in density form the core, but these models yield seemingly disparate thicknesses. Sub-nm shells have been proposed on the premise that benzene molecules could orient parallel with the silica surface, enabling strong Coulombic interactions between electrons of benzene and the polar silica surface.9 This idea is supported by NMR spectroscopy at 168 K,17 molecular dynamic simulations,9 and Raman and optical Kerr effect spectra,7 which suggest a localized shell 0.2 nm to 0.5 nm thick. Yet, 2H-NMR spectroscopy of deuterated benzene confined within SBA-15 at 110 K and 120 K places the benzene shell thickness at 2.45 nm,10 while 2H-NMR spectra of deuterated benzene confined inside porousglass mesocellular foam over the temperature range of 90 K to 180 K yields a shell thickness of 1.65 nm thickness.18 Additionally, DSC studies of confined cyclohexane place the thickness of the non-freezable surface layer at 2.14 nm.8 The SANS frozen benzene data also support a core|shell model of varied density given that the Bragg1 and Bragg2 ratios are measurably different at 153 K, Table 2. Note that it can be mathematically proven from Eqns.1 and 2 that the difference in solution density is not correlated with the presence of the porous silica, but rather this is a measure of the intrinsic benzene density as a function of radius. Moreover, in liquid form at 303 K the Bragg peaks ratios are equivalent within error bars. Our SANS data are best fit by thicker 3.5 nm shell models (for D = 12.5 nm) with a reduction in density to about 91% of bulk, Fig. 6b. Other combinations of shell thickness and density in Fig. 6b either fail to match the experimental intensity ratios for Bragg1 and Bragg2 or give incorrect ratios of √3/4 for the two Bragg peaks at the desired intensity ratio. Assuming the pore diameter is 9 nm from the N2 physisorption measurements, this places the interfacial layer thickness at ≈1.8 nm along the SiO2 walls with the remaining interfacial benzene


Page 20 of 25

Page 21 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


penetrating deep into the voids and nanoporous tunnels that inhabit the walls. However, if the pore diameter is taken to be 6.9 nm based on TEM measurements (Fig. 1b), this would place the interfacial layer thickness at only ≈0.7 nm along the SiO2 walls. It is possible the conflicting reports of shell thickness could in part reflect the extent to which different experimental techniques account for leakage into the surrounding medium. Conclusions It is shown here that binary solutions of 0.80 mole fraction benzene and 0.20 mole fraction cyclohexane confined within 9.0 nm SiO2 pores are fully miscible at 303 K, yet they clearly phase segregate below their component freezing points at 153 K. Moreover, the cyclohexane molecules preferentially segregate toward the center of the pores upon freezing as revealed by enhanced SANS sensitivity to its placement using selective deuteration/protonation of the cyclohexane. Since neither cyclohexane nor benzene chemically bond with SiO2, this segregation appears to be driven primarily by confinement. The data also indicate that upon cooling to 153 K, pure benzene becomes less dense within both the porous SiO2 and a sub-nm interfacial region located along the pore surface, while its core forms solid phase consistent with an unconfined solution. Supplemental Information Additional experimental details concerning DSC, TEM, N2 physisorption, Hg porisometry, and X-ray diffraction analysis of the mesoporous silica is available as supplemental information. Details of the SANS fits and the effect of pore wall porosity on the models are also provided.



Acknowledgements CMB and LJL are grateful to the Oklahoma Louis Stokes Alliance for Minority Participation (NSF HRD 1408748) for providing travel support. Access to NGB 30 m SANS was provided by the Center for High Resolution Neutron Scattering, a partnership between the National Institute of Standards and Technology and the National Science Foundation under Agreement No. DMR1508249. References (1)

Schreiber, A.; Ingke, K.; Findenegg, G. H. Melting and Freezing of Water in Ordered Mesoporous Silica Materials. Phys. Chem. Chem. Phys. 2001, 3, 1185-1195.

(2)

Tombari, E.; Salvetti, G.; Ferrari, C.; Johari, G. P. Thermodynamic Functions of Water and Ice Confined to 2 nm Radius Pores. J. Chem. Phys. 2005, 122, 104712.

(3)

Kittaka, S.; Ishimaru, S.; Kuranishi, M.; Matsuda, T.; Yamaguchi, T. Enthalpy and Interfacial Free Energy Changes of Water Capillary Condensed in Mesoporous Silica, MCM-41 and SBA-15. Phys. Chem. Chem. Phys. 2006, 8, 3223–3231.

(4)

Dutta, D. P.; Sharma, S. K.; Maheshwari, P.; Sudarshan, K.; Pujari, P. K. Freezing of Nanodroplets: Phase Transition of Organic Liquids Confined in Nanopores Studied by Positron Annihilation Spectroscopy. Mater. Sci. Forum 2009, 607, 218–222.

(5)

Dosseh, G.; Xia, Y.; Alba-Simionesco, C. Cyclohexane and Benzene Confined in MCM41 and SBA-15: Confinement Effects on Freezing and Melting. J. Phys. Chem. B 2003, 107, 6445–6453.

(6)

Dutta, D.; Pujari, P. K.; Sudarshan, K.; Sharma, S. K. Effect of Confinement on the Phase Transition of Benzene in Nanoporous Silica: A Positron Annihilation Study. J. Phys. Chem. C 2008, 112, 19055–19060.

(7)

Zhu, X.; Farrer, R. A.; Fourkas, J. T. Ultrafast Orientational Dynamics of Nanoconfined Benzene. J. Phys. Chem. B 2005, 109, 12724–12730.

(8)

Amanuel, S.; Bauer, H.; Bonventre, P.; Lasher, D. Nonfreezing Interfacial Layers of Cyclohexane in Nanoporous Silica. J. Phys. Chem. C 2009, 113, 18983–18986.

(9)

Coasne, B.; Fourkas, J. T. Structure and Dynamics of Benzene Confined in Silica Nanopores. J. Phys. Chem. C 2011, 115, 15471–15479.

(10)

Gedat, E.; Schreiber, A.; Albrecht, J.; Emmler, T.; Shenderovich, I.; Findenegg, G. H.; Limbach, H.-H.; Buntkowsky, G. 2 H-Solid-State NMR Study of Benzene-D 6 Confined in Mesoporous Silica SBA-15. J. Phys. Chem. B 2002, 106, 1977–1984.

(11)

Scatchard, G.; Wood, S. E.; Mochel, J. M. Vapor-Liquid Equilibrium. III. Benzene-


Page 22 of 25

Page 23 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Cyclohexane Mixtures. J. Phys. Chem. 1939, 43, 119–130. (12)

Rastogi, R. P.; Nigam, R. K. Thermodynamic Properties of Binary Mixtures of Benzene, Cyclohexane and Carbon Tetrachloride. Trans. Farad. Soc. 1959, 55, 2005–2012.

(13)

Shao, Y.; Hoang, G.; Zerda, T. W. Solid-Solid Phase Transitions of Cyclohexane in Porous Sol-Gel Glass. J. Non. Cryst. Solids 1995, 182, 309–314.

(14)

Dore, J.; Webber, B.; Strange, J.; Farman, H.; Descamps, M.; Carpentier, L. Phase Transformations for Cyclohexane in Mesoporous Silicas. Phys. A Stat. Mech. its Appl. 2004, 333, 10–16.

(15)

Mu, R.; Malhotra, V. M. Effects of Surface and Physical Confinement on the Phase Transitions of Cyclohexane in Porous Silica. Phys. Rev. B 1991, 44, 4296–4303.

(16)

Qing, Y.; Zenan, Z.; Deyu, K.; Rongshen, C.; Ethington, E. Interfacial Contact Angle Measurements of Water, Mercury, Organic Liquids on Quartz, Calcite, Montmorillonite. Constr. Build. Mater. 2007, 21, 539–545.

(17)

Xia, Y.; Dosseh, G.; Morineau, D.; Alba-simionesco, C.; Xia, Y.; Dosseh, G.; Morineau, D.; Alba-simionesco, C. Phase Diagram and Glass Transition of Confined Benzene. J. Phys. Chem. B 2006, 110, 19735–19744.

(18)

Masierak, W.; Emmler, T.; Gedat, E.; Schreiber, A.; Findenegg, G. H.; Buntkowsky, G. Microcrystallization of Benzene-d6 in Mesoporous Silica Revealed by 2H Solid-State Nuclear Magnetic Resonance. J. Phys. Chem. B 2004, 18890–18896.

(19)

Eberle, A.P.; King, H.E.; Ravikovitch, P.I.; Walters, C.C.; Rother, G.; Wesolowski, D.J. Direct Measure of the Dense Methane Phase in Gas Shale Organic Porosity by Neutron Scattering. Energy Fuels, 2016 30, 9022-9027.

(20)

Ramsay, J. D. F.; Hoinkis, E. SANS Investigations of Benzene Adsorption on Porous Silica Gel. Phys. B 1998, 248, 322–326.

(21)

Lin, M. Y.; Sinha, S. K.; Drake, J. M.; Wu, X. L.; Thiyagarajan, P.; Stanley, H. B. Study of Phase Separation of a Binary Fluid Mixture in Confined Geometry. Phys. Rev. Lett. 1994, 72, 2207.

(22)

Hellweg, T.; Schemmel, S.; Rother, G.; Brûlet, A.; Eckerlebe, H.; Findenegg, G. H. DeMixing Dynamics of a Binary Liquid System in a Controlled-Pore Glass. Eur. Phys. J. E 2003, 12, 1–4.

(23) Schemmel, S.; Rother, G.; Eckerlebe, H.; Findenegg, G. H. Local Structure of a PhaseSeparating Binary Mixture in a Mesoporous Glass Matrix Studied by Small-Angle Neutron Scattering. J. Chem. Phys. 2005, 122, 244718. (24)

Abdel Hamid, A. R.; Mhanna, R.; Lefort, R.; Ghoufi, A.; Alba-Simionesco, C.; Frick, B.; Morineau, D. Microphase Separation of Binary Liquids Confined in Cylindrical Pores. J. Phys. Chem. C 2016, 120, 9245–9252.

(25)

Kline, S. R. Reduction and Analysis of SANS and USANS Data Using Igor Pro. J. Appl. Crystal. 2006, 39, 895-900.



(26)

Jackson, C. L.; McKenna, G. B. The Melting Behavior of Organic Materials Confined in Porous Solids. J. Chem. Phys. 1990, 93, 9002-9011.

(27)

Meissner, J.; Prause, A.; Findenegg, G. H. Secondary Confinement of Water Observed in Eutectic Melting of Aqueous Salt Systems in Nanopores. J. Phys. Chem. Lett. 2016, 7, 1816–1820.

(28)

Burba, C. M.; Janzen, J. Confinement Effects on the Phase Transition Temperature of Aqueous NaCl Solutions: The Extended Gibbs-Thomson Equation. Thermochim. Acta 2015, 615, 81–87.

(29)

Burba, C. M.; Butson, E. D.; Atchley, J. R.; Johnson, M. S. Thermal Properties and Ionic Conductivities of Confined LiBF4 Dimethyl Carbonate Solutions. J. Phys. Chem. C 2014, 118, 366–375.

(30)

Sivia, D. S. Elementary Scattering Theory For X-Ray and Neutron Users; Oxford University Press: Oxford, 2011.

(31)

Guinier, A. G.; Fournet, G. Small-Angle Scattering of X-Rays; John Wiley and Sons: New York, 1955.

(32)

SASView software can be found at sasview.org.

(33)

Mildner, D. F. R.; Barker, J. G.; Kline, S. R. The Effect of Gravity on the Resolution of Small-Angle Neutron Diffraction Peaks. J. Appl. Crystal. 2011, 44, 1127–1129.

(34)

Haynes, W. M. CRC Handbook of Chemistry and Physics, 92nd Edition; CRC Press: Boca Raton, FL, 2011.

(35)

Dianoux, A.-J.; Lander, G. Neutron Data Booklet, 2nd Edition; Insitut Laue-Langevin: Grenoble, France, 2003.

(36)

Galarneau, A.; Cambon, H.; Di Renzo, F.; Ryoo, R.; Choi, M.; Fajula, F. Microporosity and Connections between Pores in SBA-15 Mesostructured Silicas as a Function of the Temperature of Synthesis. New J. Chem. 2002, 27, 73–79.

(37)

Ryoo, R.; Ko, C. H.; Kruk, M.; Antochshuk, V.; Jaroniec, M. Block-CopolymerTemplated Ordered Mesoporous Silica: Array of Uniform Mesopores or Mesopore−Micropore Network? J. Phys. Chem. B 2000, 104, 11465–11471.

(38)

Kim, S. S.; Karkamkar, A.; Pinnavaia, T. J.; Kruk, M.; Jaroniec, M. Synthesis and Characterization of Ordered, Very Large Pore MSU-H Silicas Assembled from WaterSoluble Silicates. J. Phys. Chem. B 2001, 105, 7663–7670.


Page 24 of 25

Page 25 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


TOC Graphic


Nanoconfinement-Induced Phase Segregation of Binary Benzene

Recommend Documents